帕妥珠单抗
突变
结合
药品
药物设计
计算生物学
遗传学
生物
曲妥珠单抗
药理学
生物信息学
数学
癌症
基因
数学分析
乳腺癌
作者
Xuefei Bai,Lingyi Xu,Zhe Wang,Xinlei Zhuang,Jiangtao Ning,Yanping Sun,Haibin Wang,Yugang Guo,Yingchun Xu,Jiangtao Guo,Shuqing Chen,Liqiang Pan
标识
DOI:10.1073/pnas.2413686122
摘要
Recurrent missense mutations in the human epidermal growth factor receptor 2 (HER2) have been identified across various human cancers. Among these mutations, the active S310F mutation in the HER2 extracellular domain stands out as not only oncogenic but also confers resistance to pertuzumab, an antibody drug widely used in clinical cancer therapy, by impeding its binding. In this study, we have successfully employed computational-aided rational design to undertake directed evolution of pertuzumab, resulting in the creation of an evolved pertuzumab variant named Ptz-SA. This variant, with only two mutations (T30S/D31A) located on its heavy chain, effectively reinstates binding to the mutated antigen, at the expense of a 35-fold reduction in binding affinity to HER2 (S310F) compared to the wild-type pair. Subsequently, Ptz-SA demonstrates potent killing capacity through antigen-dependent cytotoxicity. Moreover, upon engineering Ptz-SA into antibody–drug conjugates, such as Ptz-SA-MMAE, it manifests notable in vitro and in vivo antitumor efficacy by efficiently delivering cytotoxic payload into tumor cells expressing HER2 (S310F). Cryoelectron microscopy studies elucidate the molecular mechanism underlying the restored binding ability of Ptz-SA toward the S310F mutation. The steric hindrance induced by the S310F mutation is efficiently circumvented by the T30S and D31A mutations, which provides adequate space to accommodate the larger phenylalanine. Additionally, Ptz-SA also exhibits binding capacity to HER2 (S310Y), another mutation occurring at the S310 site of HER2 with high frequency. The computational-aided evolution of pertuzumab provides an alternative strategy for overcoming point mutation-mediated resistance to therapeutic antibodies.
科研通智能强力驱动
Strongly Powered by AbleSci AI